Loads fine for me.
Tell us more about how the disputes work? Seems complicated.
Here is explanation from the whitepaper, probably can't explain it better:
When the dispute is concluded:
The winning side gets the contract value from escrow, his staked rep back, and the rep of the minority voted mods.
The losing side loses only his staked rep if the freelancer and the staked rep AND the money in escrow if the employer.
The majority vote mods get proportionally to their stakes the rep of the losing side.
The minority vote mods lose their reputation at stake to the winning side.
and here is the detailed explanation
So how does it all work?
Freelancers can create listings presenting their services and push them on the network using a web client. On the backend, we use IPFS (interplanetary file system) to keep and distribute their listings on a peer-to-peer basis. No servers needed. They don’t have to keep their web client online; the offers just live in the cloud.
(Note: In the beginning, we might need to run our own nodes to boost that, while the network gains enough traction to take care of it on its own.)
Employers can search all of the listings from the same web client, or publish their own offers searchable by the freelancers.
There might be a need for insignificant fees to publish a listing in order to prevent spamming the network. It will be negligible if using the system as intended, but it will add up to non-insignificant cost if spamming. If there is a need to implement that, all of the income will be distributed proportionally to the token holders.
Once they find each other, they enter into a smart contract where:
The employer has the contract value locked in smart contract-based escrow.
Pre-agreed amount of reputation tokens are staked from each side. Each listing has a predefined amount of reputation tokens that needs to be staked in order to enter into smart contract-based agreement with. The listing creator sets that value. Both sides have to stake the same amount. The side taking the listing needs to accept that condition and stake the required amount of reputation tokens.
Predefined timeframe upon which if none of the following outcomes have been met, it self-releases everything back. That is just a safeguard to prevent funds from being indefinitely locked.
Ethearnal Reputation Tokens (ERT) can be earned in two ways:
• Directly buying them on the free market. That way, anyone can start right away and avoid the egg and chicken problem of new freelancers who need reputation to get contracts, but need contracts to get reputation.
• Taking jobs on the network and being rewarded reputation tokens upon successful completion.
In both cases, the participants have invested valuable resources (e.g. time or money) to gain that reputation, so it is valuable to them, and thus, they have initiative to keep it. Even if they don’t appreciate their time, the reputation token has monetary value on the free market.
The creator of the listing, being a gig or job offer, decides how much reputation stake to require for entering into a smart contract. So, he has the freedom to decide based on the overall contract value and his personal risk assessment.
There are three possible outcomes of this smart contract:
Both parties are happy. (They execute the contract on their own, no 3rd party intervention is needed. Both are awarded reputation tokens proportional to the contract value.)
At least one part is unhappy. (A pool of moderators steps in and decides in favor of one of the sides based on simple majority vote.)
Both are no longer willing to work together. (The contract thus self-closes after preset amount of time and returns everything in escrow and at stake.)
The tricky one is case two since only it needs involvement of a 3rd party decision. The other outcomes can be solved by the participants themselves.
OUTCOME 1:
If both parties agree on the successful execution, 99% of the contract value in escrow goes to the freelancer. The remaining 1% is used to buy rep tokens on the free market (automatically by a smart contract) at market prices, and distribute them equally to the employer and freelancer. This gives 0.5% of the value of the contract to each side in ERT tokens. This serves two purposes:
Gives to each participant reputation proportional to the value of the project.
Creates demand for the token. The demand for the token is necessary because this is fundamental for value, and token value is needed for this system to work. If there is no token value, the freelancer won't have anything valuable at stake, and will not have an initiative to act honestly.
Please note that in this case, the only "loss" of money compared to a case where participants don't use our system is 0.5%. The freelancer gets 99% of the contract value in ether and 0.5% in ERT tokens, which he can sell if he prefers money over reputation. So, he is 0.5% short. That is his cost to use a system where he is protected and can be found by employers.
The employer gets what is basically 0.5% cash back, but in ERT tokens. This is a slight incentive for them to use this system. Having more employers on the system is beneficial for the freelancers. This benefit might or might not outweigh the fee.
OUTCOME 2.
If one party is unhappy with the execution, he can open a dispute.
This action automatically opens a queue for moderators for the case. Every moderator can stake at minimum 5% reputation tokens of the predefined stake value in the job listing, and no more than 33.4%. So, effectively the employer, freelancer, and pool of moderators all have the same total reputation at stake. (Note: All moderators collectively need to have the stake amount at stake). Once enough moderators have entered the pool, so that their collectively staked ERT tokens are equal to that of the other parties, the moderation process starts. In the beginning, we will be participating in moderation as well, to ensure there are no cases with insufficient moderators. Also, 5% of the tokens that we leave for ourselves will be used for initiatives where needed, including getting moderators.
Moderator can be anyone who hold enough ERT tokens to cover the stake minimum. A stake minimum is necessary to prevent sybil attacks where someone creates lots of mods with just one ERT token each in order to dominate the moderator pool and decide in his favor. The listing creator decides how much reputation he wants at stake from the other side and the moderators pool. The minimum moderation stake is 5% from that, so the higher the stake value, the more expensive the sybil attack. The max stake of 33.4% is necessary to ensure we have at least three moderators in every case. That is needed to make the decision more objective and less likely to be manipulated.
Each moderator has a vote weight of one, regardless of how many rep tokens he stakes. If the vote was weighted proportionally to their staked rep tokens, just 2 “whale” moderators could be enough to solve any dispute.
However, what they earn from solving the dispute is proportional to their rep at stake. So, people with more rep can stake more, and naturally, can earn more since they have more at stake. This also incentivizes them to judge honestly. If they do, they gain more, and if they don’t, they lose more. On top of that, this is another initiative for people to hold tokens, since they can use them earn by moderating.
When they decide the winning side in the dispute by simple majority (51%), the rep at stake of the losing side is distributed proportionally to the moderators based on their rep at stake. Only the moderators who voted with the majority (the winning decision) get rep tokens. The mods that voted with the minority lose their stake to the winning side of the dispute. The logic is that, since they tried to rule unfairly against him, but he turned out to be right, he deserves some rep. What is fair is decided by the vote of the majority. This is necessary to give moderators an incentive to act justly. The mods who failed to vote within the determined timeframe lose their rep at stake. This will eliminate non-active mods automatically since they will be losing their rep at stake each time until there is none left and they can't be selected for mods anymore.
In case the moderator votes are at odds, only then the system looks at the tokens behind the votes to decide the majority. We can’t simply force odd number of moderators, since some of them may fail to vote, and we still get an even number.
Any feedback is welcome, we are always looking for a way to make it better.